在 2018/6/21 1:28, Jens Axboe 写道:
On 6/19/18 8:49 PM, chenxiang (M) wrote:
Hi Jens,
When i use fio to test 4k READ for 19 disks which are connected with a
expander, and find a issue:
the numbers of jobs decrease after some times, after about 2 days or 3
days, there is only one job
left, but there is no another exception. Part of log is as follows (the
attachment is the script i use to run fio
(./creat_fio_task.sh 4k read 64)):
Jobs: 19 (f=19): [RRRRRRRRRRRRRRRRRRR] [2.5% done] [7016M/0K /s]
[1713K/0 iops] [eta 02d:08h:19m:55s]
Jobs: 19 (f=19): [RRRRRRRRRRRRRRRRRRR] [2.5% done] [6978M/0K /s]
[1704K/0 iops] [eta 02d:08h:19m:55s]
Jobs: 19 (f=19): [RRRRRRRRRRRRRRRRRRR] [2.5% done] [7008M/0K /s]
[1711K/0 iops] [eta 02d:08h:19m:54s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5800M/0K /s]
[1416K/0 iops] [eta 02d:08h:19m:53s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5325M/0K /s]
[1300K/0 iops] [eta 02d:08h:19m:51s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5354M/0K /s]
[1307K/0 iops] [eta 02d:08h:19m:49s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5405M/0K /s]
[1320K/0 iops] [eta 02d:08h:19m:49s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5411M/0K /s]
[1321K/0 iops] [eta 02d:08h:19m:48s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5525M/0K /s]
[1349K/0 iops] [eta 02d:08h:19m:48s]
Jobs: 18 (f=18): [RRRR_RRRRRRRRRRRRRR] [2.5% done] [5519M/0K /s]
[1347K/0 iops] [eta 02d:08h:19m:49s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5236M/0K /s]
[1278K/0 iops] [eta 02d:08h:19m:49s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5085M/0K /s]
[1241K/0 iops] [eta 02d:08h:19m:50s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5072M/0K /s]
[1238K/0 iops] [eta 02d:08h:19m:50s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5077M/0K /s]
[1240K/0 iops] [eta 02d:08h:19m:51s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5373M/0K /s]
[1312K/0 iops] [eta 02d:08h:19m:53s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5649M/0K /s]
[1379K/0 iops] [eta 02d:08h:19m:53s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5656M/0K /s]
[1381K/0 iops] [eta 02d:08h:19m:54s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5595M/0K /s]
[1366K/0 iops] [eta 02d:08h:19m:54s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4995M/0K /s]
[1219K/0 iops] [eta 02d:08h:19m:54s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4976M/0K /s]
[1215K/0 iops] [eta 02d:08h:19m:52s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4996M/0K /s]
[1220K/0 iops] [eta 02d:08h:19m:51s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4980M/0K /s]
[1216K/0 iops] [eta 02d:08h:19m:50s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4990M/0K /s]
[1218K/0 iops] [eta 02d:08h:19m:51s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5007M/0K /s]
[1222K/0 iops] [eta 02d:08h:19m:52s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5012M/0K /s]
[1224K/0 iops] [eta 02d:08h:19m:53s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [5075M/0K /s]
[1239K/0 iops] [eta 02d:08h:19m:54s]
Jobs: 17 (f=17): [RR_R_RRRRRRRRRRRRRR] [2.5% done] [4994M/0K /s]
[1219K/0 iops] [eta 02d:08h:19m:55s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4893M/0K /s]
[1195K/0 iops] [eta 02d:08h:20m:34s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4620M/0K /s]
[1128K/0 iops] [eta 02d:08h:20m:35s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4600M/0K /s]
[1123K/0 iops] [eta 02d:08h:20m:35s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4595M/0K /s]
[1122K/0 iops] [eta 02d:08h:20m:34s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4600M/0K /s]
[1123K/0 iops] [eta 02d:08h:20m:33s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4603M/0K /s]
[1124K/0 iops] [eta 02d:08h:20m:32s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4625M/0K /s]
[1129K/0 iops] [eta 02d:08h:20m:31s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4638M/0K /s]
[1132K/0 iops] [eta 02d:08h:20m:29s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4626M/0K /s]
[1129K/0 iops] [eta 02d:08h:20m:27s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4616M/0K /s]
[1127K/0 iops] [eta 02d:08h:20m:26s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4615M/0K /s]
[1127K/0 iops] [eta 02d:08h:20m:25s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4624M/0K /s]
[1129K/0 iops] [eta 02d:08h:20m:25s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4634M/0K /s]
[1131K/0 iops] [eta 02d:08h:20m:25s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4635M/0K /s]
[1132K/0 iops] [eta 02d:08h:20m:25s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4626M/0K /s]
[1129K/0 iops] [eta 02d:08h:20m:25s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4614M/0K /s]
[1127K/0 iops] [eta 02d:08h:20m:26s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4606M/0K /s]
[1125K/0 iops] [eta 02d:08h:20m:27s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4595M/0K /s]
[1122K/0 iops] [eta 02d:08h:20m:28s]
Jobs: 16 (f=16): [RR_R_RR_RRRRRRRRRRR] [2.5% done] [4631M/0K /s]
[1131K/0 iops] [eta 02d:08h:20m:28s]
......
Is it normal or exception like that? If exception, do you know why?
The jobs are probably finishing. You set a runtime, but didn't make
it time based. If you add time_based=1 it might work closer to
what you expect.
Ok, thanks.
I'm trimming the CC list, you added every linux list except the
list you should have added, which is the fio list...
I didn't notice there was a linux list for fio before
--
To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
